The aims of the proposed research are to critically evaluate, compare and validate selected multi-modality imaging metrics as quantitative (surrogate) biomarkers of the response of breast tumors to specific treatments to provide the scientific basis for their translation into patient management and clinical trials. Recent years have seen a dramatic increase in the range of information available from imaging methods so that a number of techniques are available to quantitatively monitor tumor growth and treatment response. Several of these have been used in both pre-clinical and clinical studies, but with mixed results confounded by lack of standardization, inadequate understanding of underlying mechanisms and absence of appropriate validation to assist their interpretation. We propose to systematically evaluate emerging, clinically-viable imaging metrics in appropriate animal models to establish which combination of methods is most accurate at predicting response to specific treatments that are relevant for breast cancer. The paradigm we have chosen to achieve these ends considers two major categories of human breast cancer, HER2 positive tumors and ER/PR/HER2 triple-negative tumors and both established and emerging therapies. For each category we will assess treatment response by performing longitudinal studies that combine PET, SPECT, and MRI to provide functional assessments of the response of breast tumors to treatment. We will also correlate imaging with histology data to understand the mechanistic underpinnings of the information provided by each type of measurement. We hypothesize that treatment type will determine which imaging metrics are most sensitive to early response. To test this hypothesis we will pursue the following specific aims: 1. [HER2+ cancer] In the BT-474 mouse model of breast cancer with and without resistance to Herceptin (to simulate responders and nonresponders, respectively), measure the effects of treatment response to Herceptin or Herceptin+lapatinib as reported by PET, SPECT, and MRI metrics. This study will also be performed in the polymoma middle T (PyMT) spontaneous mouse model of human breast cancer. 2. [Triple negative cancer] In the MDA-231 mouse model of triple-negative breast cancer with and without resistance to Docetaxel (to simulate responders and nonresponders, respectively), measure the effects of treatment to Docetaxel or Docetaxel+sunitinib as reported by PET, SPECT, and MRI metrics. PUBLIC HEALTH RELEVANCE: We propose to systematically evaluate emerging clinically-viable imaging metrics to establish which methods are most accurate at predicting treatment response with clinically employed breast cancer treatments. In this way we hope to expand the range of quantitative imaging biomarkers that can be implemented in the clinical setting. We hypothesize that treatment type will determine which imaging metrics are the most sensitive to early response enabling specific treatment classes to be paired with specific imaging approaches so that the most sensitive imaging biomarkers can be selected when designing clinical trials.